My name is James (GeoidDude), and I am part of the Apple Data Team (#adt), along with Andrew and Luke (ruggs) who have posted here in the past. I am reaching out, because our team has noticed a complicated situation regarding some sections of motorway in Kuala Lumpur, which appear to be accessible to pedestrians. I am curious about the community’s opinion regarding some details.
Is it common for sidewalks to exist along motorways in Kuala Lumpur or throughout Malaysia? Our team has noticed them on portions of the E10 and E11, but could there be others?
My primary question relates to the highway=motorway classification which implies foot=no. This conflicts with the examples I shared above. There is a section in the highway=motorway description on the Malaysian Roads Tagging Wiki which states, “Consider using highway=trunk for certain stretch of the highway that features lack of controlled access.” Would these highways be more appropriately classified as trunks? I understand this would be a pretty drastic change, and would require a significant editing effort, but the walkability of these highways seems important to consider in OSM, because of the pedestrian infrastructure that exists along these highways. I believe some other options to account for pedestrian access at these locations would involve adding sidewalk=* and foot=yes tags to motorway segments, or adding nearby highway=footway geometry.
Does this group have a preferred solution to this inconsistency? Our team would love to help improve this aspect of the data, and would appreciate any opinions on the matter.
Well, that is tolled road, which is why it is originally tagged as a motorway I think. No, in general our motorways don’t allow pedestrians. In big cities however, some of the motorways were of previously upgraded main road that cuts through mixed use areas, hence the inevitablility of having sidewalks on certain sections of them to minimize construction cost (rather than having a totally isolated motorway). In addition, I believe that local mappers rarely map pedestrian related features since they are more focused on getting the car navigation part complete first. It just picked up until recently when pedestrian routing (plus public transport) started to show up on google. Even then, many are actually contributed by non-locals.
My preference is either make it all tolled trunk road, or keep it as it is (as a special case with your additional recommendation to add highway footway). Both these options doesn’t seem to have any impact on routing softwares, and I do prefer your solution. Having a highway with jumbled sections of classification doesn’t seem right to me.
Only the second example looks like it might still be a sidewalk.
In the first example, I doubt that you can get from the pedestrian bridge to the sidewalk.
In the third example, the sidewalk is separated from the carriage-way by a fence. That it is not much different from the German motorway bridge over the Rhine river between Wiesbaden and Mainz (it’s a one-meter high concrete barrier here: https://www.openstreetmap.org/way/328265861#map=16/50.0361/8.2117 ) - thus it is a separate way, and not part of the motorway.
Thanks for the responses! It seems like the preferred solution is to map these pedestrian features using separate highway=footway geometry. Depending on the situation, a footway=sidewalk tag may also be appropriate. I will wait a few more days for more responses before my team proceeds with any edits. Thanks for helping to clarify the situation.